IMPORTS AND DATA EXPLORATION¶

In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
In [2]:
sns.set_theme(style = 'darkgrid')
In [3]:
data = pd.read_csv("./data/CAR DETAILS FROM CAR DEKHO.csv")
data.head()
Out[3]:
name year selling_price km_driven fuel seller_type transmission owner
0 Maruti 800 AC 2007 60000 70000 Petrol Individual Manual First Owner
1 Maruti Wagon R LXI Minor 2007 135000 50000 Petrol Individual Manual First Owner
2 Hyundai Verna 1.6 SX 2012 600000 100000 Diesel Individual Manual First Owner
3 Datsun RediGO T Option 2017 250000 46000 Petrol Individual Manual First Owner
4 Honda Amaze VX i-DTEC 2014 450000 141000 Diesel Individual Manual Second Owner
In [4]:
data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4340 entries, 0 to 4339
Data columns (total 8 columns):
 #   Column         Non-Null Count  Dtype 
---  ------         --------------  ----- 
 0   name           4340 non-null   object
 1   year           4340 non-null   int64 
 2   selling_price  4340 non-null   int64 
 3   km_driven      4340 non-null   int64 
 4   fuel           4340 non-null   object
 5   seller_type    4340 non-null   object
 6   transmission   4340 non-null   object
 7   owner          4340 non-null   object
dtypes: int64(3), object(5)
memory usage: 271.4+ KB
In [5]:
data.shape
Out[5]:
(4340, 8)
In [6]:
data.isnull().sum()
Out[6]:
name             0
year             0
selling_price    0
km_driven        0
fuel             0
seller_type      0
transmission     0
owner            0
dtype: int64
In [7]:
data.columns
Out[7]:
Index(['name', 'year', 'selling_price', 'km_driven', 'fuel', 'seller_type',
       'transmission', 'owner'],
      dtype='object')
In [8]:
data['make'] = data['name'].apply(lambda x: ' '.join(x.split(' ')[:1]))
data.head()
Out[8]:
name year selling_price km_driven fuel seller_type transmission owner make
0 Maruti 800 AC 2007 60000 70000 Petrol Individual Manual First Owner Maruti
1 Maruti Wagon R LXI Minor 2007 135000 50000 Petrol Individual Manual First Owner Maruti
2 Hyundai Verna 1.6 SX 2012 600000 100000 Diesel Individual Manual First Owner Hyundai
3 Datsun RediGO T Option 2017 250000 46000 Petrol Individual Manual First Owner Datsun
4 Honda Amaze VX i-DTEC 2014 450000 141000 Diesel Individual Manual Second Owner Honda

VISUALIZATIONS¶

H.BAR - TOP 10 MAKES AND MODELS¶

In [9]:
make_val = data.make.value_counts()[:10]
df1 = pd.DataFrame()
df1['make'] = make_val.index
df1['count'] = make_val.values
df1
Out[9]:
make count
0 Maruti 1280
1 Hyundai 821
2 Mahindra 365
3 Tata 361
4 Honda 252
5 Ford 238
6 Toyota 206
7 Chevrolet 188
8 Renault 146
9 Volkswagen 107
In [10]:
fig = px.bar(
    df1,
    x = 'make',
    y = 'count',
    title = 'VALUE COUNTS OF TOP 10 CAR MANUFACTURERS'
)
fig.show()

We see that Maruti Suzuki is by far the most popular among the cars sold in this period of time, followed by Hyundai and close competition between Mahindra and Tata. This goes on to show that Maruti Cars are greatly resold and have good scope for reselling.

COUNTPLOT - YEAR WISE COUNT OF PRODUCED CARS¶

In [11]:
year_cnt = pd.DataFrame()
year_cnt['year'] = data.year.value_counts().index
year_cnt['count'] = data.year.value_counts().values
year_cnt
Out[11]:
year count
0 2017 466
1 2015 421
2 2012 415
3 2013 386
4 2014 367
5 2018 366
6 2016 357
7 2011 271
8 2010 234
9 2019 195
10 2009 193
11 2008 145
12 2007 134
13 2006 110
14 2005 85
15 2020 48
16 2004 42
17 2003 23
18 2002 21
19 2001 20
20 1998 12
21 2000 12
22 1999 10
23 1997 3
24 1996 2
25 1995 1
26 1992 1
In [12]:
fig = px.histogram(
    data,
    x = 'year',
    title = "YEAR WISE COUNT OF CARS PRODUCED"
    
)
fig.show()

Going by the countplot, we see that in this period of time, reselling cars have peaked around the period of 2013 to 2018. There is a sharp decline in 2019 and 2020, posibbly because of the COVID 19 debacle.

PIE CHART - DEALER TYPE COUNTS¶

In [13]:
seller_data = pd.DataFrame()
seller_data['type'] = data.seller_type.value_counts().index
seller_data['count'] = data.seller_type.value_counts().values
seller_data
Out[13]:
type count
0 Individual 3244
1 Dealer 994
2 Trustmark Dealer 102
In [14]:
fig = px.pie(
    seller_data,
    values = 'count',
    names = 'type',
    title = "COUNTS OF SELLER TYPES"
)
fig.show()

This pie chart shows that most of the resale is happening through individual sellers. Maybe this can be attributed to the rise of apps like CarDekho and Carwale, along with classics like OLX.

PIE CHART - SHARE OF TRANSMISSION TYPE¶

In [15]:
plt.figure(figsize = (14, 7))
labels = ['Automatic', 'Manual']
plt.pie(
    data['transmission'].value_counts().sort_values(),
    labels = labels,
    explode = [0.1,0.1],
    autopct='%1.2f%%',
    colors = ['lightgray','cyan'], 
    startangle = 90
)
plt.title('SHARE OF TRANSMISSION TYPE')
plt.axis('equal')
plt.show()

This chart is to demonstrate that a huge majority of the resold cars are of manual transmission type. Possibly because automatic cars have better life due to less wear and tear on the working of the transmission.

VIOLIN PLOT - DISTRIBUTION OF TRANSMISSION TYPE OF CARS BASED ON FUEL TYPE¶

In [16]:
plt.figure(figsize = (13, 9))
sns.violinplot(
    data = data, 
    x = 'year', 
    y = 'fuel',
    hue = 'transmission'
)
plt.show()

Thought this plot shows a lot of information, one interesting fact is that even though CNG fuel type is lesser known and even lesser posessed, it has the same distribution of Petrol cars sold.

One more fact is that the general distribution of Diesel cars being resold is smaller than petrol ones, again showing that diesel cars have better life.

DISTPLOT - DISTRIBUTION OF DRIVEN KILOMETERS¶

In [17]:
plt.figure(figsize = (13, 8))
sns.kdeplot(
    data = data,
    x = 'km_driven',
    fill = True,
    hue = 'transmission',
    palette = 'flare'
)
plt.show()

This simple looking plot uncovers one crucial and daily experienced inference : The mean of the kms driven of resold cars lies around the 80 - 120k region. This is hugely due to the fact that most of these are normal consumer cars, and that the tinkering and modification industry in India is very limited and niche, thus resulting in people not holding their cars for long, before moving on.

HEATMAP - CORRELATION BETWEEN FEATURES¶

In [18]:
df1 = data[['year', 'selling_price', 'km_driven']]

corr = df1.corr()

fig, ax = plt.subplots(figsize=(8, 6))

sns.heatmap(
    corr,
    center = 0, 
    square = True,
    annot = True
)

plt.title("CORRELATION HEATMAP")
plt.show()

The main inference from this heatmap is that there is a medium positive correlation between Selling Price and year. This is natural, that the inflation has caused the selling price of cars to increase along the years.

SCATTERPLOT - RELATIONSHIP BETWEEN SELLING PRICE AND KILOMETERS DRIVEN¶

In [19]:
plt.figure(figsize = (14, 8))
fig = px.scatter(
    data,
    x = 'selling_price',
    y = 'km_driven',
    title = 'RELATIONSHIP BETWEEN SELLING PRICE AND KMS DRIVEN'
)
fig.show()

<Figure size 1008x576 with 0 Axes>

This scatterplot demonstrates the common inference: The selling price decreases as the kms driven are increased. But it is not as linear as expected, due to some people selling their cars for lower price even though they are not used much, maybe due to urgent need for cash.